Conventional digital communications utilize the intensity of a signal pulse to encode a binary bit of information, using a full or high amplitude pulse to convey the 1 bit of information and a low or no amplitude pulse to signal the 0 bit datum, or vice versa. Wireless communications operate under the constraints of finite bandwidth and multiple sources of signal interference that can cause high bit error rate levels in data streams using simple digital codes. Spread-spectrum signaling protocols have been developed to address these issues in a manner that provide meaningful data rates and higher signal integrity. These protocols utilize phase shift keying (PSK) or quadrature amplitude modulation techniques to shape the transmitted pulse into a symbol that encodes a series of consecutive data bits into a single pulse. FIG. 1A uses a phase map to depict how a signal with constant amplitude 101 and modulated phase shift 103 produces four different phase states 105A, 105B, 105C, 105D that are used to symbolize four different two-bit combinations of data [1,1], [0,1], [1,0], [0,0]. FIG. 1B depicts how a signal constellation containing 16 different amplitude and phase states 107 is used to encode four (4) bits per symbol: [1,1,1,1], [1,1,1,0], [1,1,0,1], [1,0,1,1], [0,1,1], [1,0,0,1], [1,0,0,0], [0,1,0,1], [0,1,0,0], [0,0,1,0], [1,0,0,1], [0,0,0,1], [1,0,0,0], [0,0,0,1], [0,1,1,0], and [0,0,0,0].
These symbol modulation techniques affect the shape of the pulse through the signal roll-off parameter, α, which can take on values ranging between 1≦α≦0. As shown in FIGS. 2A, 2B, 2C, different roll-off parameters used to represent different symbols will not significantly modulate the signal carrier when the pulse 109A, 109B, 109C is viewed in the time domain. Time domain signal modulation primarily affects the leading and trailing tails 111A, 111A′, 111B, 111B′, 111C, 111C′ of the pulse. FIGS. 3A, 3B, 3C depicts how the varying roll-off parameters affect the pulse in the frequency domain 113A, 113B, 113C. T is used to define the symbol time length and W is the Nyquist rate, W=½T, in FIGS. 2, 3. This modulation format causes the pulse's power spectral density to be spread over more frequencies as the roll-off parameter is increased. Therefore, symbol detection is more efficiently performed by analyzing the symbol in the frequency domain.
Conventional receivers will use sensors to register the pulse's time domain signature and dedicate processor functions to perform inverse Fast Fourier Transforms (IFFF) or inverse Discrete Fast Fourier Transforms that mathematically compute pulse's power spectral density. The use of mathematical methods to de-convolve a symbol's power spectral density adds component cost to the receiver, consumes additional power from any available power budget, and occupies valuable real estate when packaged on a mobile wireless platform.
This is particularly so in wireless communications systems based on orthogonal frequency division multiplexing (OFDM). OFDM techniques, including, but not limited to, WiMAX systems, are multi-carrier modulation methods developed to boost data rates reliably. Boosting data rates on a single carrier by shortening symbol time lengths is often more susceptible to increased bit error rates. OFDM methods use a plurality of carriers (often referred to as “sub-carriers”) operating at a lower data rate. This allows the composite data of all sub-carriers to be communicated at a combined rate that is comparable to the data rate of a single carrier with the same channel bandwidth using the same basic modulation at a higher data rate. The principal advantages to using longer symbol duration times is a net reduction in error rate by reduced susceptibility to errors from inter-symbol interference caused by multi-path time dispersion. Furthermore, inter-symbol interference is not as problematic when frequency-selective fading is distributed only over a few of the sub-carriers and the fading depth over the majority of sub-carriers is not great enough to generate significant bit errors. FIGS. 4A, 4B show how a plurality of sub-carriers 115 produce a composite power spectra of the individual sub-carriers that has a rectangular shape with most of the modulated data contained in the power spectral densities of leading/trailing side bands 117A, 117A′. However, the greater number of sub-carriers also increases the IFFF and/or IDFFF computational processing power needed to interpret the sub-carrier symbols, which processing power is limited on a mobile platform.
Passive resistor, capacitor, and inductor components, collectively referred to as passive components, are used to form the filtering stages that comprise these devices. At present, most passive components are assembled on the surface of a printed circuit board and generally comprise 80% of all the components used in a fully populated circuit board and account for 50% of the real estate occupied on the circuit board's major surface. Small form factor is a general requirement for mobile wireless systems. Therefore, methods that reduce a circuit's footprint by transferring the passive components from the circuit board's surface to one or more interior layers are desirable. This practice, more commonly known as embedded passive technology, is also useful in larger scale high-speed circuits, such as servers and telecommunications switches, which require a large number of electrically terminated transmission lines. Although embedded passive technologies have been under development since the early 1980's, very few approaches have been successful in meeting optimal performance tolerances. The inability to rework (exchange) an out-of-tolerance passive component after it has been embedded into the circuit board requires tolerances of ±1% of targeted performance for these components, since the failure of a single component causes the entire board value to be lost. Additionally, it is desirable for embedded passives to maintain their targeted performance tolerances over all anticipated operating temperatures to facilitate design and ensure circuit reliability. Operating temperatures typically range from −40° C. to +125° C. Most of the prior art on embedded passive technologies relies on thin film technologies that comprise a layer of material with uniform dielectric properties. Resistive metal thin films have demonstrated the greatest ability to achieve thermally stable performance with tolerances within ±1%.
As shown in FIGS. 5A, 5B, embedded resistors 119 typically comprise two metal sheets consisting of a conductive metal layer 121 and a resistive metal layer 123 that are affixed to respective dielectric layers 124, 125. The respective metal layers are patterned to produce conductive leads 127 within the conductive metal layer 121 and resistor elements 129, 130 in resistive metal layer 123. This results in the location of resistor elements 129, 130 over gaps 131, 132 between conductive leads 127 when the two laminated sheets are aligned and brought into contact as depicted in FIG. 5B. The resistance of the resistor elements, 129, 131 is controlled by the spacing of gaps 131, 132 between conductive leads 127, and the sheet resistivity and thickness of the resistive metal layer 123. Elemental resistance is determined by the spacing of gaps 131, 132 as the laminated resistive metal sheet will have uniform thickness and resistivity. In general, dimensional controls of the thin film metallic resistor elements 129, 130 will achieve tolerances of ±5% and laser trimming is used to bring the performance tolerance to within ±1% of the targeted value. Resistive thin films comprised of nickel (Ni) or platinum (Pt) have low thermal coefficients of resistance (TCR) that provide thermal stability within a tolerance of ±1% over operating temperatures. Primary reliability issues with thin film resistor elements 129, 130 include interactions with metallic electrodes 127 and/or the material forming dielectric layer 125 that cause metallic plaques to form, as well as mismatches between the thermal coefficients of expansion that may cause delamination between the thin film resistor elements 129, 130 and encapsulating dielectric 125.
Thin film techniques are also used to fabricate embedded capacitors. Demand for embedded capacitors has been driven largely by a need to suppress power noise in high-speed CMOS semiconductor circuits, wherein simultaneous flipping of switching devices draws a large surge current that is supplied by the power plane embedded in the circuit board to which the semiconductor device is attached. Power noise is generated in the circuit when inadequate charge is available from within the power plane to supply the surge current. Decoupling capacitors are used to suppress power noise in high-speed circuits. FIGS. 6A, 6B generally depict the use of embedded decoupling capacitors 131 to suppress power noise in a circuit comprising a circuit board 133 and a high-speed semiconductor chip 135. The power plane 137 is constructed as a laminate sheet capacitor consisting of a dielectric layer 139 inserted between two conducting metal sheets 141, 142 (FIG. 6A). One of the conducting metal layers 142 is patterned to provide floating ground planes 143 (FIG. 6B) in areas where the power plane 137 maintains electrical contact with a via 145 that supplies surge currents to the semiconductor device. Embedded decoupling capacitors 131, formed by dielectric layer 139 being located between the metal sheet 141 and the floating ground plane 143, collect and supply surplus charge to suppress power noise that can be created when there is inadequate stored charge to supply the surge current. C-Ply material manufactured by 3M Company utilizes a dielectric layer 139 that consists of an epoxy loaded with high-K dielectric barium titanate powders. While these structures provide high sheet capacitance (6 nF/inch2), higher capacitance values are desired. Furthermore, the dielectric powders loaded into the dielectric layer 139 have grain sizes (1-2 micron) that are inadequate to provide a low thermal coefficient of capacitance (TCC) and stable temperature performance. As such, 3M C-Ply decoupling capacitors are only rated to have X7R-type behavior (performance values within ±15% of the targeted value over anticipating operating temperatures). Fujitsu, Ltd. has reported the development of an aerosol deposition technique that forms a film of high-κ barium titanate ceramic at room temperature, which can be used to embed capacitors within a circuit board using an aerosol of pre-formulated ceramic powders. Grain sizes reported to be required to form high-quality films (0.05-2 micron) are also insufficient to maintain a TCC suitable for stable thermal performance or COG-type behavior.
FIGS. 7A, 7B show top and profile views of a passive inductor 150 embedded into an interconnect structure or printed circuit board 151. Inductor 150 typically comprises a coil structure 147, or a simple loop structure (not shown), configured on one or more dielectric sheets 152, with a feed line 149 supplying current to the coil 147, that is situated on a separate dielectric layer 153 located inferior (or superior) to the plane(s) upon which the coil is located. The dielectric layers 152, 153, 154 are generally comprised of the identical material used to construct the circuit.
U.S. Pat. No. 5,154,973 to Imagawa, et al., disclose a dielectric lens antenna that includes high-κ dielectric ceramic compositions prepared from powders with a mean particle size ranging between 1 and 50 micron that are mixed with an organic thermoplastic material. U.S. Pat. No. 5,892,489 to Kanba et al., disclose a chip antenna incorporating high-κ oxide ceramics formed from powders having a mean particle size of 10 microns. U.S. patent No. to K-D Koo, et al., disclose a chip antenna which comprises helically wound conductors formed by printing planar trace structures on dielectric sheets and assembling those sheets to form said chip antenna. U.S. Pat. No. 6,028,568 to K. Asakura, et al., disclose a chip antenna containing at least one folded antenna formed by printing conductor on a plurality of dielectric layers, wherein at least one dielectric layer is a magnetic material, and fusing said dielectric layers into a solid structure. U.S. Pat. No. 6,222,489 B1 to T. Tsuru, et al., disclose a chip antenna containing monopole or dipole antenna prepared by printing conductor traces on a plurality of dielectric layers, wherein each individual layer has uniform composition providing said individual layer with either a relative permittivity of ∈R=1-130 or relative permeability of μR 1-7, and said plurality of layers is fused into a single component. U.S. Pat. No. 6,650,303 B2 to H. J. Kim, et al., disclose a chip antenna comprising a plurality of dielectric sheets, wherein each sheet has uniform composition throughout the sheet, and conductor leads that are configured to form a helical antenna. U.S. Pat. Nos. 6,680,700 B2 and 6,683,576 B2 to A. Hilgers disclose a chip antenna that comprises a core ceramic substrate having uniform dielectric properties and conducting metal traces on its periphery that is surface mounted to a circuit board. U.S. Ser. No. 6,025,811 to Canora et al. disclose a directional antenna with a closely-coupled director embedded or mounted on a circuit wherein the director element is a conductive element that has a rectangular cross-sectional profile. U.S. Ser. No. 10/265,351 filed by T. T. Kodas et al. disclose inkjet techniques to form conductive electronic materials from a colloidal suspension of nanoparticles in a low viscosity solvent. U.S. Ser. No. 10/286,363 filed by Koda et al. disclose direct-write (syringe-based) methods to form inorganic resistors and capacitors from a flowable high-viscosity precursor solution consisting of combination of molecular precursors and inorganic powders. U.S. Pat. Nos. 6,036,899 and 5,882,722 disclose methods and formulations to apply metallization layers using nano-particle pastes.
U.S. Pat. No. 6,027,826 to de Rochemont, et al., disclose articles and methods to form oxide ceramic on metal substrates to form laminate, filament and wire metal-ceramic composite structures using metalorganic (molecular) precursor solutions and liquid aerosol spray techniques. U.S. Pat. Nos. 6,323,549 and 6,742,249 to de Rochemont, et al., disclose articles that comprise, and methods to construct, an interconnect structure that electrically contacts a semiconductor chip to a larger system using at least on discrete wire that is embedded in silica ceramic, as well as methods to embed passive components within said interconnect structure using metalorganic (molecular) precursor solutions and liquid aerosol spray techniques. U.S. Pat. Nos. 5,707,715 and 6,143,432 to de Rochemont, et al., disclose articles and methods to relieve thermally-induced mechanical stress in metal-ceramic circuit boards and metal-ceramic and ceramic-ceramic composite structures prepared from a solution of metalorganic (molecular) precursors, and further discloses the incorporation of secondary phase particles (powders) in said solution of said solution of metalorganic (molecular) precursors. The contents of each of these references are incorporated herein by reference as if laid out in their entirety. U.S. Ser. No. 11/243,422 discloses articles and methods to impart frequency selectivity and thermal stability to a miniaturized antenna element, and the construction of simplified RF front-end architectures in a single ceramic module.